ARCHITECTURE: MATERIAL AND I M A G I N E D
92
The Use of Building Technology in Cultural
Forensics: A Pre-Columbian Case Study
M. IVER WAHL, AIA
University of Oklahoma
INTRODUCTION
This paper poses the possibility that as early as 750 AD,
technology transfer may have taken place between PreColumbian Peru and the Anasazi culture in UtaWColorado.
The possibility of such a cultural link became apparent to
this researcher in 1976 while executing graduate field work
in Central America; however, further study has awaited
adequate funding. If such transfer existed, the probable route
would seem to have been from Peru north along the Pacific
coast, then up the Colorado River to the San Juan area of
UtaWColorado.
PURPOSE
The first objective of this study has been to determine if this
tentative hypothesis is sufficiently plausible to warrant
further testing. It is a status report of preliminary research
work that is currently in progress. The second objective of
this study was to explore the possible usefulness of architectural technology as an overt and objective measure in
evaluating such plausibility. This study specifically focused
on an attempt to track the evolution of a reasonably unique
roof structure in both the upper Mantaro river area near
Ayacucho, Peru and in the San Juan drainage area in Utah/
Colorado.
Because of the great distances involved, even if similarities in building technology can be adequately identified.
additional architectural similarities will need to be sought in
subsequent studies before a final hypothesis can be defined
that will warrant testing. Further, if such a cultural link
existed, there should be corroborating evidence among other
indicators traditionally used by archaeologists such as pottery, linguistics, and other primitive artifacts.
BACKGROUND
For decades, archaeologists and anthropologists have had
and interest in establishing links between Pre-Columbian
Mesoamerica (Mexico and Central America), and the cultures of the American Southwest. The "kiva" in the "Four-
Corners" area ofthe American Southwest is perhaps the most
unique building form of the Anasazi culture, but it does not
seem to exist in Mesoamerica. In the world of architecture,
this argues significantly against significant Mesoamerican/
Anasazi ''diffusion" theories.
The prevailing theory for the "in-situ" evolution of the
kiva building type was proposed by John Otis Brew in 1946.
Brew dismissed outside cultural diffusion as necessary to
kiva development based on his field studies at Alkali Ridge,
Utah. Briefly his theory traces the shift from the Basket I11
pit house (used as a domestic residence), to the ceremonial
kiva configuration common to Pueblo I1 period of the
Anasazi culture. His study of this subject is thorough, and
continues to largely accepted by archaeologists. However,
this well-developed theory did not specifically address the
unique "corbeled-dome" or "cribbed roof structure used at
many Mesa VerdeIHovenweep Anasazi sites. From an
architect's viewpoint it is arguably one of the greatest
innovations in building technology offered by the native
American experience.
ASSUMPTIONS
In pursuing this draft hypothesis. it was necessary to accept
the following assumptions:
1. That by their education and practice architects have a
significant role to play in cultural forensics involving the
interpretation of ancient architectural technology.
Webster defines "forensics" as public debate in support
of subsequent judicial proceedings, or the use of professional skill to produce evidence in a judicial proceeding
establishing legal responsibility. In this current project
this legal analogy gives rise to the potential use of other
judicial concepts such as motive, opportunity, and levels
of proof that can be reasonably required.
2. That cultural judgments being pursued through public
debate necessarily involve "surmise." Surmise is defined
by A. J. Jaffe as judgment made in a realm of uncertainty,
and holds that it is justified in theories involving migration of ancient cultures.
85TH
ACSA A N N U A L M E E T I N G ANDTECHNOLOGY CONFERENCE
3. That the circular or D-shaped kiva of the Mesa Verdel
Hovenweep Anasazi has sufficient similarities in fonn
and construction to represent an identifiable architectural "type."
93
4. That the circular kiva form should be recognizable in
aerial photos.
5. That the corbeled-dome kiva roof is sufficiently unique to
be considered a cultural "signature" item for the Anasazi.
AII('H1TECTIIRE: MATERIAL A N D IMAGINED
Fig. 3 , Plan view of an Anasazi "cribbed or corbeled dome" kiva
roof structure.
greatest interest is from 500 AD through 1300 AD. Where
justified, limited study has been given to issues outside this
titne frame.
PROBLEMS AND LIMITATIONS
Fig. 2. A typical Mesa VerdeiHovenwcep Anasazi kiva. Note thc
perimeter bench and pilasters.
SCOPE
This paper is based on site examinations in the American
Southwest conducted well before 1976, site reconnaissance
and analysis completed during graduate work in several
countries of Mesoamerica ( 1976), sabbatical study conducted in residence at selected Anasazi sites (between 1987
and 1995), and reconnaissance completed at selected PreColumbian sites in Peru ( 1995).
The sites of greatest interest in the United States are
Hovenweep, Mesa Verde. Aztec. and Chaco Canyon. Pisac,
Machu Picchu, Sacsayhuainan, Wari, Tunasniyoq. Huaca
Pucllana, and the National Museu~nof Anthropology and
Archaeology in Lima, were studied in Peru.
Up to this time the study has only involved one researcher.
Sites in the United States Southwest. Mexico. Central, and
South America have been studied. Thc time period of
Surmise: Jaffe states that appropriate rigor in using surmise
requires first that the researcher has adequate knowledge of
the subject matter, and second that all available data must be
reviewed. With respect to the first requirement some
adequacy in knowledge of architecture can be claimed
however, in the issues of linguistics, pottery, and other areas
of archaeology, no such claim of adequacy can be made.
Since at this point only a plausible hypothesis is sought, not
final proof, the second requirement has only been partially
met.
Barriers to (/ifusion: Thornas Mails in his book "Pueblo
Children of Mother Earth" argues against diffusion between
cultures that are very distant from each other since an
external idea must undergo a selective process before being
integrated into a new, and potentially very different culture.
Kiva Types: Watson Smith in his book "When is a kiva"
emphasizes the difficulty in defining a "typical" kiva given
the considerable diversity of kiva design and construction.
Overcotni17g/CornplevlerltingAccepted Kiva . Theory:
Brew's theory of in-situ development for the kiva is so
convincing that any new hypothesis addressing the evolution of the kiva must either have the potential to bring
ovenvhelming evidence to deny this well-accepted interpretation. or it must be compatible with it.
,4 single architectlrrnl feature or construction method
{naynot be sufficient by itselfto assure sufficient plausibility
for further testing of a new hypothesis. For this reason,
corroboration in other architectural aspects and by other
acceptcd archaeological ~ncthodssuch as pottery was sought.
85THACSAANNUAL MEETING ANDTECHNOLOGY CONFERENCE
METHOD
Phase One - Literary Reseurch began in 1976 and has
continued through 1996 on an intennittent basis. Available
aerial photographs were also studied. Evidence found in
these aerial photographs first suggested the hypothesis. and
guided the planning of the 1995 site reconnaissance effort in
Peru
Phase TWO- Foi.n~alSite Dafu Collection was conducted
at selected sites in the San Juan drainage area in Utah,'
Colorado, and at selected sites in Peru. Data collected during
earlier research at other sites in the Southwest, Mexico. and
Central America was also utilized. Some disaster site
reconnaissance methods were employed.
Phase Three - Datu Atzaly.~isbegan prior to leaving each
site, but was pri~narilycompleted in residence within the
United States. It is primarily a comparative study and makes
use of checklist. matrices, and overlays following a methods
developed by this researcher while completing graduate
work at the University of Colorado, and utilized in thesis
work for Kansas State University during field data collection
in Central America in 1976.
SUMMARY OF DATA
Selected Anasazi Sites showed considerable similarity of
building footprint, wall construction, and the use of attached
pilasters to support a corbeled-dome roof structure. The
Anasazi kiva of Pueblo I1 period and later consistently had
a religiousicerelnonial use. At most sites this innovative roof
structure appeared fairly suddenly without evidence of local
evolution. With the exception of the Alkali Ridge excavation it has had the appearance either a technical import, or
as a leap of separate intuitive insight. Brew's interpretation
does not address the structural need or revolutionary long
span potential of this roof structure. From an architectural
standpoint, this structural innovation is not needed at Alkali
ridge due the very short spans involved. The largest was
about nine meters and could easily be spanned by other
systems that were already in use at the site.
Selected Inca sites showed some similarity in building
footprint. contradictory functional building use in some
cases, and no similarity in roof construction. For example.
an elaborate water supply system is an important part of the
hnction of the Muyuc Marca at Sacsahuaman. At Pisac and
Machu Picchu the Intiwatanta had no roof, no pilasters, and
was completely filled with a large single stone used in sun
worship.
Selected Proto-Incan Sites: Only the Vegachayoq Moqo
complex at Wari (or Huari) nearthe modern townofAyacucho
showed significant similarity in building footprint, apparent
hnctional use, and wall construction to an Anasazi kiva. A
four foot wide bench occupies about one third of the perimeter of this D-shaped ceremonial space. Pilasters interrupt
the bench areas at a regular interval (about 10 feet). Most
importantly these pilasters may have supported a cribbed
roof covering a portion of the perimeter of this ceremonial
Fig. 4. The Muyuc Marca at Sacsahuaman near Cuzco, Peru
space. Unforhmately little excavation has taken place at this
site other than the 1953 expedition by Bennett and the 1982
excavation by Bragayrac.
ANALYSIS
Architectural Similarities
The apparent function of the kivas of the Mesa Verdel
Hovenweep Anasazi and the selected building at Wari were
both ceremonial. Attached pilasters in association with a
bench surrounding a ceremonial space is found in Anasazi
sites, and the same condition exists for a portion of the
perimeter at Wari. This is a critical feature in both Anasazi
and Peruvian archaeology since stone evidence is better
preserved than the wooden roof system. Unfortunately, the
actual pattern of roof structures have occasionally been
ignored during excavation, or were found to be burned. A
proto-cribbed or corbeled structure seems reasonable given
the geometry of the supports and climate at Wari. Such a roof
might represent a transitional step to the h l l cribbed or
corbeled-dome structure that is the focus of this paper. The
geographic distance from Peru to Utah argues strongly
against its involvement as a precedent for the American
southwest, but architecturally the substantial span at Huari
(70 feet across the ceremonial space) verses the much shorter
span (27 feet maximum) at Alkali Ridge, Utah at least makes
the need for such a long span structure a more plausible at
Wari.
The climate and farming methods are strikingly similar
between the Huarpai Huaril Chanka cultures of Peru, and the
Mesa VerdeIHovenweep Anasazi in ColoradoIUtah. NorthSouth orientation of the ceremonial rooms in both the Upper
Mantaro and the San Juan drainage are quite common. The
placement of round kivas and their large size is similar
between the early North Coast cultures in Peru and that
found at Aztec National Monument in New Mexico. The
topography is quite similar in the Upper Mantaro and the San
Juan drainage area. These ceremonial spaces were often
located on the mesa tops in both areas.
The masonry wall construction, to include rough stone
.\R('HITECTL!KE: MATERIAL A N D IMAGINED
96
Fig. 5. Vegachayoq Moqo at Wari. Peru. Note the D-Shaped footprint.
shaping and adobe plastering, is colnmon to both the Upper
Mantaro and the San Juan cultures. Thicker walls occasionally receive rough fill between coursed masonry in both
cultural areas. Lintels that are almost identical in construction can be found at both Huari in Peru, and Hovenweep in
UtahlColorado.
Fig. 6 . Plan of the potential "proto-kiva" at Wari, extracted from
field notes by Wahl in 1995.
Other Apparent Similarities
Hopi legends of their migration tell of coming by ocean from
Asia. This is possible given existing currents. It further
describes some groups proceeding south, and some going
north along the Pacific coast. To go north would have been
easier by water if the traveler stays close to the coast. This
route would lead up the Sea of Cortez to the mouth of the
Colorado river. In its day, the Colorado river and its
tributaries were the superhighways offering food, shelter,
and water in an arid land. Successive waves of travelers have
evidently used this same route. The Hopis call those that
migrated through the Aleutians the "back-door people."
The Zuni language is considered a linguistic isolate.
Links to other languages has proven problematic and controversial. No complete Zuni dictionary is apparently available; however, enticing similarities seem to exist. According to the local Quechuan speakers at Pisac and Machu
Picchu the central urban feature of these important cities is
the6'anchoror center for the sun." In the Quechuan language
of Peru the "intiwatana." According to Alvin M. Josephy Jr,
the Zuni word for their town is "itiwana" meaning "the center
(of their world)." Great diversity of language roots exists in
85T"
ACSA A N N U A L M E E T I N G ANDTECHNOLOGY CONFERENCE
97
Straight lines that are apparent in Aerial Photographs at
Chaco Canyon appear to be similar to the forty-one ceques
at Cuzco, Inca roads which may have been builtby the ~ " a r i ,
and to some of the Nazca lines on the south coast of Peru.
Ball courts ofthe Mesoamerican type are found at Snaketown
in the American Southwest. This is unusual since this is a
characteristic of the east coast of Mexico and the Maya in
Central America.
To the untutored eye of this researcher, there are some
similarities in the form and decoration of pottery of the
Ayacucho area in Peru, and Pueblo I1 pottery found at
Anasazi sites such as Alkali Ridge and Mesa Verde. Boats
sufficient for extended maritime travel existed in Peru as
early as 600 AD. A unique tule (reed) boat is still used by
traditional individual fisherman along the north coast of
Peru. This same type of reed boat is also constructed by
Paiute Indians of the American Southwest. This boat is an
excellent design for the ocean, but is not as u s e h l on desrt
lakes as other primitive reed boats built elsewhere in the
ancient world.
CONCLUSIONS
Motive: Potential motives for use of the Pacific coast route
Fig. 7. View of the context for Wari, Peru. Note the similarity to
the American Southwest.
the native Americans of the southwest. This is particularly
true of the pueblo Indians at Hopi, Zuni, Acoma, and the Rio
Grande Pueblos of New Mexico. In San Juan county Brian
Stubbs has demonstrated linguistic ties to Hopi. Aztec,
Serrano. Luisno, O'odham. North and South Tepehuan,
Yaki. Mayo, Tarahumara, Ute, and Nulnic among others.
The importance of this set of linguistic groups is that they
geographically line the route for technology transfer suggested in this paper.
Both the Anasazi and the cultures of Peru had a "vertical
economy." It has been demonstrated that with wide variation
in altitude, climate varies substantially. This was very
important to both agricultural and hunterlgatherer societies
since each zone inherently had a different food producing
potential. Trade and seasonal migration across these elevation changes seems to have been an important feature in both
Peru and the American Southwest.
Parrots are present in the pueblos of the American Southwest as early as 1100 AD. Hopi legends hold that they are
a clan that migrated from far to the south, parrot petroglyphs
appear at Hovenweep in Utah, and parrots play a significant
role in current Hopi ritual. Parrots of the type that has been
found by archaeologist in the Southwest are from at least as
far south as southern Mcxico. Trade is therefore believed to
have come through Casa Grande in Chihuahua. Mexico.
Fig. 8. Typical wall and lintel construction at Wari.
ARC H I T E C T O K E M A T E R I A L A N D I M A G I N E D
of transfer would have varied widely. Trade, immigration
due to a lack of success at their current site (such as the parrot
clan of Hopi legend), escape from progressive integration
and consolidation of power in Peru, and escape from the
numerous wars and economic conflicts that characterize the
period from 500 to 1500 AD would be the most likely. For
those escaping wars or religious oppression in Peru, they
could not head south since one of the most arid deserts in the
world is there. The Andes mountains and the powerhl
opposition of Tiahuanaco and the Cuzco Inca were to the
west. Nazca on the coast and Huari were connected by trade
and excellent roads. This route was about a 10 day walk.
Opportunity: The time of expansion and consolidation in
the Peruvian highlands coincides with the development of
the roof system in question in the American Southwest. The
shortest estimate of the time required to migrate from Alaska
to Tierra del Fuego is about 1000 to 2000 years. Once the
route was known and populated it should have been faster
going the other direction for trade or continued migration as
outlined in Hopi and Zuni legend. That leaves about 12000
years for such contact to develop prior to the roof structure
in question. In law this lnight be considered "opportunity"
in a temporal sense.
Means: Existing currents and adequate boats existed to
exploit the potential water route up the Pacific coast. There
is a growing body of evidence that at least portions of this
route were heavily used for trade.
Resolution with existing theories: Brew's in-situ evolution interpretation remains convincing. That outside ideas
may have accompanied relatively small groups of migrating
people across the San Juan area remains possible, and can
coexist with Brew's interpretation. Ideas from such individuals or groups would face selective acceptance before
being integrated into local society. People from a similar
climate, topography, and agricultural tradition might possess such ideas. It is widely recognized that maize spread
from one end of the New World to the other. Therefore, at
least agricultural technology has been proven convincingly
to have existed along the Pacific coast as asserted in the
proposed hypothesis. The existence of the Paiutes tule boats
in allnost identical fonn to that in use on the north coast of
Peru suggests strongly that maritime technology may also
have made the trip.
Building Technology as Evidence: It was fortunate that
the roof being investigated was relatively unique. Comparisons of specific building construction methods proved to be
significantly more objective than other judgments of "kiva"
similarities in this study. Supporting archaeological reports
were very important in establishing the probable internal
hnctions for the buildings studied. However, surmise of
significant proportion was required in interpreting the internal functions of those buildings for which adequate archaeological reports were not available. For this reason. greater
confidence in judgments relating to specific building technology is justified.
Unfortunately the actual roof configuration at Huari still
remains in question. This, and the fragmentary nature of the
evidence at this time leads me to conclude that insufficient
architectural weight is brought to the argument for the
plausibility of the working hypothesis at this time. More
architectural research and analysis is both needed and scheduled for the spring of 1997 in Peru before the hypothesis can
be finalized for testing.
ACKNOWLEDGEMENTS
Funding for the work has come, in part, from a University of
Oklahoma sabbatical. The generous cooperation of Dr. Ellen
Foppes and other personnel at Mesa VerdelHovenweep National
Monument has been essential. Consultation and literary sources
provided by Dr. Cirilo Vivanco Pomacanchari of the Social
Science Faculty at the University of Huamanga lead to the inclusion of the Chanka culture in this study. His continued participation is anticipated in the next stage of this project.
REFERENCES
Bennett, Wendell C. E,~cavationsat Wari, Avacucho, Peru. Yale
University Publications in Anthropology Vol. 49. New Haven:
Yale University Press. 1953. 1-14 and 41-45
Bragayrac, Enrique. Archaeological Excavations in the Vegachayoq
Moqo Sector ofHuari. Huari Administrative Structure, Prehistoric Monumental Architecture, and State Government. Edited
by William H. lsabell andGordan G. McEwan. Washington,DC:
Dunbarton Oaks Research Library Collection , 1991. 54-57
Brew, John 0 . Archaeology of Alkali Ridge, Southeastern Utah.
Papers of the Peabody Museum of American Archaeology and
Ethnology Vol. 21. Cambridge: Harvard University Museum,
1943. 152-2 14 and 273
Coe, Michael, Dean Snow, and Elizabeth Benson. Atlas ofAncient
Atnerica. New York: Facts on File Publications, 1986. 8-9 and
134-191
Hemming, John. Motlunzents of the liicas. Boston: Little, Brown,
and Co. 1982. 134-141
Isabell, William H. Prehistoric Grotmd Drawings of Peru. PreColumbian Archaeology. San Francisco: W. F. Freeman and
CO., 1980 188- 196
Isabell. William H. State Origins in the AJtacucho k l l e v , Central
Highlands. Peru. The Origins and Development of the Andean
States. Edited by Jonathan Haas, Shelia Pozorski. and Thomas
Pozorski. Cambridge: Cambridge Univ. Press, 1987. 83-96
Jaffee, A .J. The First In~niigrantsfi-otnAsia. New York: Plenum
Press, 1992. 12-65
Mails, Thomas E. Tlie Pueblo Children of Mother Earth Vol. 1.
Garden City: Doubleday and Co. Inc.. 1983. 4-5
Patterson, Alex. Field Guide to Rock Art Sjwbols of the Greater
Southwest. Boulder: Johnson Books, 1992. 152- 153
Ruhlen, Merritt. A Guide to the CVorld's Languages Vol. I .
Stanford: Stanford University Press, 1987. 200-21 8
Sasser, Elizabeth S. Architecture ofAncient Per-~i.Lubbock: 1969.
Smith. Watson. When is a Kivn, and Other Questions About
So~ithwe.~t
Arch(~eologj~.
Tucson: University of Arizona Press,
1990. 54-57
Stubbs. Watson. The Narive Languages of;Snn J~mnCounty. Blue
Mountain Shadows Vol. 13. Blanding: San Juan County Historical Commission. 1994. 63-67
Waters. Frank. TAc Book of Hopi. New York: Penguin Books.
1963. 1-58
Wheat. Margaret M. S ~ i ~ ~ Arts
~ i l ~ofthe
n l Prin~itivePaiutes. Reno:
University of Nebada Press. 1967. 40-46
Young. Jane M. Slgtlsfi.o~nthe Ancestors. Albuquerque: Univcrsity of Ncw Mexico Press. 1988. 96-101